This competing renewal application, by a multidisciplinary group of investigators with expertise in cancer proteomics and cancer prevention, is focused on the application of proteomics to the identification of tumor antigens that induce a humoral response in subjects with cancer. There is evidence that a humoral response to such protein antigens may predate the diagnosis of cancer, indicating the utility of assays of specific circulating tumor protein antigens or their corresponding autoantibodies, for the early detection of cancer. The proposal builds on accomplishments during the current funding period in which a proteomics approach was utilized that consisted of two-dimensional gel electrophoresis to separate tumor cell and tumor tissue proteins followed by Western blotting and incubation of individual blots with a subject serum. Proteins that reacted specifically with antibodies in sera from newly diagnosed patients with a particular type of cancer relative to controls were identified by mass spectrometry. The tumor types investigated to date have included lung, colon and pancreas as well as breast, liver, esophagus and neuroblastoma, resulting in the identification of multiple antigens/autoantibodies that have potential utility as markers for early cancer diagnosis. Some of these markers have been selected by EDRN for validation studies. In this renewal application, we propose to focus on three types of adenocarcinomas: lung, colon and pancreas to expand our current panel of diagnostic antigens detectable in circulation and/or for which subjects with cancer have circulating auto-antibodies. In specific aim 1, we propose to replace the two-dimensional gel approach with an innovative microarray based high-throughput discovery platform developed by our group, in which natural proteins from tumors and cell lines are separated and spotted on microarrays. Microarrays containing several thousand spots, which can be prepared in large numbers, are incubated each with a subject serum. Spotted proteins that react specifically with sera from a particular cancer are identified by mass spectrometry. The goal is to establish, for each of the three tumor types, a robust panel of markers that has utility for early cancer diagnosis. In specific aim 2, we propose to undertake detailed studies of marker antigens identified to date and additional markers identified in specific aim 1, to determine the basis for their antigenicity. Studies will include assessment of the role of post-translational modification and elucidation of the particular antigenic epitopes. We also propose to develop assays for marker antigens that would be utilized for their independent validation by EDRN. A network for the procurement of serum and tissues from subjects with cancer or with preneoplastic lesions to achieve the goals of this project has been put into place by the applicant group with credits provided to referring centers, through an NCI approved protocol. This Biomarkers Development Laboratory will substantially contribute to the goal of developing and validating biomarkers for early cancer detection through its own resources and through an EDRN collaborative network it has established.